This report is based on an analytical study of migration patterns in Ukraine following the outbreak of the full-scale invasion on 24 February 2022. It is based on an analysis of aggregated, anonymised data on the movements of Vodafone mobile subscribers, information from the State Voter Register, and the OPORA Civil Network’s own calculations.

Detailed information, interactive data visualisations and the full data from this study are available on a dedicated dashboard linked here: https://oporaua.org/drv

Key findings

Overall mobility and the true scale

  • 8.3% of Vodafone subscribers changed their place of residence within the country after 24 February 2022.
  • The figures below are merely a conservative estimate (lower bound), as movements involving currently inactive SIM cards or an unknown destination (see methodology) account for the majority of all migration patterns studied; therefore, the actual extent of mobility may be several times higher.

Internal migration

  • 1.2 million people changed their region of residence within Ukraine.
  • A further 640,000 people moved within their own region, changing districts.
  • 482 unique migration routes between regions were recorded.
  • The Kyiv Region, the city of Kyiv and the Lviv Region have seen the highest population growth. In contrast, the eastern regions of Ukraine are experiencing the greatest outflow of migrants.

Outward migration

  • Around 5.2 million people may have left the country between 2022 and 2025.
  • Outward migration peaked in 2022, with 2.4 million people leaving Ukraine that year, accounting for 46.2% of the total number of migrants during the analysed period.

Discrepancies with the State Voter Register

  • The State Voter Register (hereinafter the Register) reflects only around 42% of inter-regional mobility (0.5 million official address changes recorded, compared with 1.2 million actual moves).

  • The Register data reflects the arrival of only around 45,000 citizens in the territory of the overseas constituency (OC) — compared to around 5.2 million migrants abroad, as indicated by Vodafone data. However, the Register also records departures from the OC, so the net migration balance for the constituency is negative (-14,700). Thus, the electoral address of almost all Ukrainians abroad does not match their actual place of residence.
  • Migration trends (inflow or outflow), as recorded by the Register and the mobile operator Vodafone, coincide in only 14 out of 22 regions.

Methodology

Study parameters and data source

Data from Vodafone Ukraine, a mobile network operator, was used for the calculations.

Time frame: 23 February 2022 to 1 March 2026

Reference dates: as at 23 February 2022 and as at 1 March 2026.

Priority breakdowns: gender, age, social status.

Additionally, the mobile operator’s share of the total number of subscribers for the period was calculated for further extrapolation.

Objective of the study

The objective of the study was to quantitatively assess and identify trends in population movement following 24 February 2022, as well as to evaluate the feasibility of updating the State Voter Register to reflect the actual distribution of Ukrainian citizens in the context of the migration crisis caused by Russia’s armed aggression.

Key definitions

Permanent place of residence (PPR) refers to the subscriber’s home location as defined by the operator’s methodology, calculated on the basis of the last three full months prior to the reference date. To compare periods separated by time, the operator uses only aggregated residence data.

Internal migration is recorded as a change of district or region between two points in time, 23 February 2022 and 1 March 2026; the intermediate trajectory is not tracked.

Outward migration is examined by year. The region of departure is considered to be the subscriber’s last place of residence immediately prior to departure, rather than their pre-war permanent place of residence. Departure to the territory of Russia and Belarus is also classified as outward migration. However, moving to temporarily occupied territory is not classified as outward migration, nor is departure from occupied territory to government-controlled territory. The outward migration indicator reflects specifically the crossing of the state border, rather than forced displacement within the internationally recognised territory of Ukraine. OPORA analysed this parameter in order to identify trends and quantitative indicators of outward migration, rather than to determine the countries of destination for migrants.

Breakdown categories

Gender: male, female, unknown.

Age groups: 0–17, 18–34, 35–49, 50–69, 70+, unknown.

Social status: schoolchild, university student, employee, self-employed individual, pensioner, other, unknown.

Method for calculating absolute values

For calculations within each region, the average Vodafone Ukraine penetration rate in the region and the SIM card penetration rate per capita were used.

Formula: Vodafone data × (1 ÷ penetration rate) ÷ number of SIM cards per capita

As Vodafone covers only part of the market, absolute values were obtained by extrapolation: the findings of the analysis were extrapolated to the subscriber base of all operators on the assumption that subscribers of other operators behave similarly to Vodafone subscribers.

Restrictions on data access and the areas excluded from the study

The regions of Donetsk, Luhansk and Kherson, as well as the Autonomous Republic of Crimea and the city of Sevastopol, have been excluded from the analysis of internal migration at the request of the operator’s security service. These restrictions did not apply to outward migration, so these regions are included in the out-migration data but not in the internal migration data.

In addition, areas bordering Russia and Belarus, as well as parts of the Zaporizhzhia and Kharkiv Regions, have been excluded from the study. This interactive map illustrates the areas excluded from the analysis, together with the temporarily occupied territory (according to DeepState Map) for comparison.

If a subscriber resided in an area falling within the scope of the analysis and subsequently moved to a permanent place of residence in an area excluded from the study, they were classified under the “unknown location” category (see the explanation below for further details).

The “unknown location” category

One of the key methodological caveats concerns subscribers whose area and district of arrival during relocation are undefined. This category may include:

  • subscribers who have moved abroad and deactivated their SIM card;
  • subscribers who have remained in temporarily occupied territory;
  • subscribers who have switched to another operator;
  • subscribers with very low network activity;
  • subscribers who have moved permanently to an area subject to security restrictions.

Unfortunately, movements with an unknown destination account for the majority of all recorded migrations.

The “unknown location” category is not synonymous with relocation, as a subscriber may have remained in the same place but simply opted out of the operator’s services. Furthermore, a person who has purchased a new SIM card (even from the same operator) is classified as a new user, and their migration is not tracked in the manner prescribed by the methodology of this study.

The “unknown location” category may encompass various situations, such as actual relocation (outward migration, moving to occupied or restricted territories), discontinuing the use of an operator’s services without changing one’s place of residence, or a significant reduction in the frequency of using an operator’s services. Consequently, these data cannot simply be added to migration figures, nor can they be completely disregarded. Therefore, records with an unknown location have been excluded from the study, and findings regarding departures abroad have been presented separately.

The study covers a smaller proportion of recorded relocation cases — those in which the destination could be identified. The figures cited in the report represent a conservative estimate of confirmed movements. The actual scale of mobility and discrepancies with the State Population Register data significantly exceed these absolute figures.

Interpretation limitations

The following interpretation limitations should be borne in mind when analysing the data:

Two point-in-time snapshots rather than a trajectory. Internal migration is measured as the difference between the two dates: 23 February 2022 and 1 March 2026; interim movements, returns and multi-stage routes are not reflected.

Renaming of districts. Due to the administrative reform, the Civil Network OPORA uses only the renamed (decolonised) names of the administrative units (Chervonohrad District → Sheptytskyi District, Krasnohrad District → Berestyn District, Novomoskovsk District → Samar District, etc.).

Break in data coverage. For occupied territories, the loss of mobile connectivity manifests as “zero” figures in later years, which does not imply that migration has ceased.

Impact of the extrapolation hypothesis. Absolute values are indicative. Rather than specific figures, general trends and proportions should be considered the most reliable.

When comparing with the Register, the different nature of the sources should be taken into account. Since mobile data records actual location, whereas the State Voter Register records the legal address for voting, these sources measure different things. Accordingly, discrepancies between them are to be expected and quite meaningful.

OPORA’s previous research

To better understand the nature of the migration waves triggered by targeted attacks on the energy sector and critical infrastructure, please refer to OPORA’s other analytical reports:

How Russia’s targeted missile strikes on Kyiv in May 2023 affected the migration patterns of its residents. A joint study by the OPORA Civil Network, the International Centre for Ukrainian Victory and Kyivstar (2023) reveals how, between 15 April and 15 June 2023, some Kyiv residents left the city to seek temporary respite from the relentless terror. However, the availability of sufficient air defence resources proved to be a decisive factor in averting a critical situation and preventing a panic-stricken mass exodus of residents that could have been irreversible.

How Russia’s missile strikes of Kyiv in October 2022 and January 2023 affected the migration patterns of its residents. A study by the OPORA Civil Network, the International Centre for Ukrainian Victory and Vodafone (2024) analyses the impact of Russian shelling on population migration in Ukraine between October 2022 and January 2023. The report establishes a clear causal link between large-scale missile strikes on energy infrastructure and peaks in the number of Ukrainians leaving the country.

Challenges

More than four years after the start of Russia’s full-scale invasion, Ukraine is facing the worst demographic crisis in Europe since the Second World War. According to various estimates, the number of Ukrainians who have been forced to flee abroad could range from 4,000,000 to 7,000,000, with the vast majority now living in Europe.

Meanwhile, according to the Ministry of Social Policy of Ukraine, as of 31 May 2026, there were approximately 4.6 million officially registered internally displaced persons (IDPs) within the country. Although the official estimate may currently be somewhat inflated or fail to accurately reflect the scale of displacement, it bears witness to an unprecedented migration of Ukrainians.

However, the State Voter Register (the Register), which de facto represents the institutional record of the population’s distribution and serves as the foundation for organising any election, is in fact based on pre-war migration data. As of 30 April 2026, the Register recorded 34,136,031 voters, of whom 1,457,790 (4.3%) have no registered address, and around 7,940,000 (22%) are formally registered in the temporarily occupied territories, where the Ukrainian state is unable to carry out any real verification of voters.

According to the Central Election Commission, the Register contains data on 42,000 voters in the temporarily occupied territories aged over 100 and over 1,000 voters aged over 115. These figures highlight the need to verify the register’s data.

It is quite clear that without a thorough update of the Register data, active voter registration abroad, and coordination with IDP registers and the State Border Guard Service’s Border Crossing Register, the first post-war elections risk taking place with a predictably low actual turnout. The critical discrepancy between the official voter lists and reality goes far beyond the margin of error permitted by international observation standards, which will inevitably undermine the overall legitimacy of the voting results.

This study offers one of the first systematic quantitative assessments of the discrepancies between the official data of the State Voter Register and actual voter migration since 2022. Similar studies utilising other types of objective data will facilitate proper, timely and systematic verification of the Register.

Internal migration

The study analysed the movements of over 10.8 million Vodafone subscribers who were within the area covered by the analysis as of 23 February 2022 and 1 March 2026. The findings show that 8.3% of them changed their place of residence within Ukraine.

The calculations have been extrapolated to the subscriber base of all mobile operators, resulting in 1.2 million inter-regional and 640,000 intra-regional movements. These figures are a conservative estimate, and the actual scale of internal migration is most likely higher, as for methodological reasons the areas with the largest population outflow (the regions of Donetsk, Luhansk, Kherson, and parts of the Kharkiv and Zaporizhzhia Regions) were excluded from the calculation.

The net migration balance is positive in 10 regions of Ukraine and negative in 12.

Regions with the largest population growth: Kyiv city (+58,000), Kyiv Region (34,500), Lviv Region (32,000) and Zakarpattia Region (28,000).

The largest population outflow was recorded in the regions of Kharkiv (-102,000), Zaporizhzhia (-28,000) and Mykolaiv (-22,000).

In relative terms, the highest growth rates were recorded in the regions of Zakarpattia (+154%) and Chernivtsi (+100%) (Growth here is calculated as the ratio of the net migration balance to outflow: (Inflow − Outflow) ÷ Outflow × 100%. In other words, +100% means that twice as many people arrived as left).

As methodological limitations preclude the possibility of capturing population outflow from the Donetsk, Luhansk and Kherson Regions, the figure of “-102,000” for the Kharkiv Region represents only the “tip of the iceberg” of the outflow.

The five most popular routes account for around 20% of all movements. The two dominant routes are “opposite” ones within the capital’s agglomeration: Kyiv → Kyiv Region (97,000) and Kyiv Region → Kyiv (87,000).

Notably, 7 out of the 10 leading routes are linked to Kyiv, confirming its role as the main national “magnet”. Routes from the Kharkiv Region make up the second largest cluster, serving as one of the evacuation corridors from the frontline eastern communities.

Kyiv is the predominant destination from all neighbouring regions: the capital city accounts for 59.4% of all departures from the Kyiv Region, 38.3% from the Chernihiv Region, 30.3% from the Zhytomyr Region, 25.9% from the Cherkasy Region, and 23.6% from the Poltava Region.

In addition, we can identify “regional migration hubs” in the west and south: 33.5% of migrants from the Ivano-Frankivsk Region choose the Lviv Region, whilst 33.5% of departures from the Mykolaiv Region are to the Odesa Region.

The high proportion of movements between neighbouring regions may also indicate low long-distance mobility for a significant portion of the population.

Ukraine’s migration network is extremely uneven: in most regions, there is a single dominant direction accounting for 20–30% of relocations.

Among intra-regional movements, the Lviv Region stands out as having the most, accounting for over 22% of all intra-regional movements in the country (or 142,000 people). The high level of migration activity in the Kharkiv Region (15%) and the Dnipropetrovsk Region (13.1%) may indicate intense population turnover within the large frontline regions.

The demographic profile of IDPs indicates that the distribution of migrants by gender is almost equal: 49.8% are women, 47.3% are men, and 2.8% are of unknown gender.

In terms of social status, wage earners are the largest group, accounting for 70%. Next come university students (11.9%), pensioners (8.6%), self-employed individuals (5.1%), schoolchildren (3.2%) and others (1.2%).

Over 75% of displaced persons fall within the economically active population; young people account for 15.1%.

Even though the main reason for displacement is Russia’s full-scale invasion, the economically active population dominates among IDPs — meaning that relocation is not only a way to escape danger, but also an adaptive response by people of working age.

Outward migration

The study’s findings indicate that around 5.2 million Ukrainians moved abroad between 2022 and 2025, with almost half of these departures occurring in 2022 (2.4 million people). It is currently impossible to determine exactly how many of those who migrated have returned — for example, with a different SIM card from Vodafone or another mobile operator. Nevertheless, the available data suggests that following the peak in 2022, annual outflows stabilised at 0.86–0.99 million people per year, although migration from Ukraine has not ceased, having entered a protracted, ongoing phase.

Year

Departures, thou.

Percentage of total

Notes

2022

2,393.4

46.2%

Peak year

2023

932

18%

Decline following a peak

2024

992

19.2%

Slight upward trend

2025

860

16.6%

Continuation of the downward trend

The largest number of people left the country from the Lviv Region (763,000) — likely due to its geographical proximity to the EU border. It should be noted that those leaving the region may also have included IDPs who had previously moved to the Lviv Region from other parts of Ukraine and for whom the region served merely as a temporary refuge.

The regions with the highest rates of outward migration are listed below.

Region

2022

2023

2024

2025

Total, thou.

Lviv Region

296.1

173.3

163.7

129.9

763

city of Kyiv

107.1

75

94.5

60.3

336.9

Kharkiv Region

143.7

47.4

58

51.7

300.8

Odesa

97.4

72

70.2

59.7

299.3

Vinnytsia Region

135.8

49.8

61.2

47

293.8

Donetsk Region

227.6

30.7

22.2

5

285.5

Zakarpattia Region

117.6

44.9

46.8

65.1

274.4

Dnipropetrovsk

120.6

47.5

47.3

50.9

266.3

Khmelnytskyi Region

99.1

41.8

54.1

34.4

229.4

Ternopil Region

81

53.1

45.8

35.8

215.7

Luhansk Region

197.6

–

–

–

197.6

A total of 285,500 departures from the Donetsk Region were recorded over the entire period, with 80% of these (227,600) occurring in 2022.

In contrast, there were 197,600 departures from the Luhansk Region in 2022, but data for subsequent years are no longer available. These figures are not indicative of a halt in migration itself, but rather reflect the actual security situation, hostilities and the inability to monitor subscriber movements following the occupation.

It should be noted that the inclusion of these regions in statistics on out-migration, but their absence from the internal migration data, is explained by a number of methodological limitations: security restrictions do not apply to data on departures abroad.

Demographic structure of migrants

The economically active population of working age (the 35–49 age group) predominantly makes up the number of migrants abroad. Employees account for 77.1% of the total number of migrants abroad, which is even higher than among internal migrants (70%).

The full breakdown by social status (according to Vodafone’s methodology) and a comparison with internal migration are shown in the table below.

Category

Migrants abroad

Internally displaced persons

Employed workers

77.1%

70%

No data

14.2%

-

University student

4%

11.9%

Self-employed individual

1.7%

5.1%

Schoolchild

1.5%

3.2%

Pensioner

1.1%

8.6%

Other

0.4%

1.2%

A comparison with internal migration reveals a clear structural difference: more vulnerable age and social groups are significantly more likely to move within the country than abroad:

  • Proportion of pensioners among migrants abroad is only 1.1%, compared with 8.6% among internal migrants;
  • Proportion of university students is 4% compared to 11.9%;
  • Proportion of schoolchildren is 1.5% compared to 3.2%;
  • Proportion of self-employed individuals is 1.7% compared to 5.1%.

The economically active population (employed workers + self-employed individuals) represents 78.8% of migrants abroad, compared to 75.1% of internal migrants; young people (university students + schoolchildren) account for 5.5% compared to 15.1%. Thus, most of those who leave the country are of working age, whilst more vulnerable groups tend to move within the country.

It is worth noting that women account for around 78% of the total number of pensioners who leave the country. This imbalance is due both to objective demographic factors (higher average life expectancy among women) and to certain restrictions on border crossings for men of conscription age (23–60 y.o.).

Official changes of residence (Register)

According to the State Voter Register, as of mid-January 2026:

  • 1.9 million people had changed their electoral address;
  • 1.4 million (73.7%) had moved within their own region;
  • 500,000 had moved to another region.

In terms of the volume of intra-regional movements, leading regions are Dnipropetrovsk, Vinnytsia and Lviv. The highest number of voter migrations between regions was recorded among voters in the city of Kyiv, the overseas constituency and the Kyiv Region.

The following key indicators stand out:

  • The Kyiv Region recorded the largest official increase in voters (+28,900 or +26% in terms of the dynamics of outflow/inflow).
  • The Autonomous Republic of Crimea, Sevastopol, and the regions of Kherson and Luhansk show a trend of −100%, meaning the Register records zero arrivals but shows departures;
  • The overseas constituency has a negative balance of −14,700 voters, despite figures in the millions for actual migrants according to Vodafone. According to the Register, only 44,900 persons officially arrived in the overseas constituency, meaning departures from the constituency exceed arrivals by approximately 14,700;
  • The highest volume of internal migration within a single region was recorded in the Dnipropetrovsk Region. The section of the population that formally updates their electoral address does so predominantly locally (within the region);
  • Six out of ten leading “official” routes are linked to an overseas constituency. This indicates significant movement in this direction, with almost no official changes to electoral addresses taking place.

The primary “official” routes for changing one’s electoral address:

Route

Number of people

Percentage of total, %

city of Kyiv → Kyiv Region

24,848

5%

Kyiv Region → city of Kyiv

14,659

2.9%

OC → Lviv Region

7,227

1.4%

OC → Zakarpattia Region

6,189

1.2%

OC → Ivano-Frankivsk Region

5,391

1.1%

city of Kyiv → Zhytomyr Region

5,235

1%

OC → city of Kyiv

5,119

1%

city of Kyiv → Cherkasy Region

4,818

1%

Zakarpattia Region → OC

4,595

0.9%

city of Kyiv → OC

4,585

0.9%

Taken together, the 10 most active migration routes account for just 16.4% of the total number of officially registered moves. The bulk of migration is spread across more than 480 pairs of regions, with figures ranging from a few hundred to several thousand persons. Furthermore, the share of the most common route (the city of Kyiv → Kyiv Region) accounts for only 5% of the total volume.

Thus, the mobility officially recorded in the Register is highly fragmented: migration processes are not concentrated along a limited number of movement patterns, but are decentralised and represented by hundreds of local flows between different regions. This may provide further support for the argument that the Register poorly reflects actual movements: there are indeed regions in Ukraine that have become hubs for IDPs, but these hubs are not reflected in the official routes for changing one’s electoral address.

Comparison of the Vodafone and Register data

This section compares two sources of data on population movements: data from the mobile operator Vodafone (actual location of subscribers) and the State Voter Register (registered address). Before analysing the discrepancies, it is important to note a key difference in the nature of how this data is measured: they record movements with slightly different “depths”.

  • According to the study methodology, as reported by Vodafone, a move is recorded only when a subscriber crosses the boundary of a district (leaves the boundaries of their previous district of permanent residence). A change of address within the same settlement or the same district is not classified under any of the categories of relocation.
  • The Register, however, records any change of electoral address, including moves within the same settlement: moving between neighbouring houses is formally classified as a change of electoral address.

Consequently, two approaches to recording the same fact can be observed: the “Vodafone indicator” refers to actual movements between districts and beyond, whilst the “Register indicator” covers legal changes of address of any scale, including the most minor ones. This should be borne in mind when comparing absolute figures.

Regional comparison of net migration
The net migration figures reported by the Register and Vodafone matched in 14 out of 22 regions (64%): 9 regions show an inflow according to both sources, whilst 5 show an outflow. In the remaining 8 regions, the population movement trends identified by the Register and Vodafone contradict each other: the city of Kyiv, and the regions of Volyn, Cherkasy, Zhytomyr, Ternopil, Chernihiv, Khmelnytskyi and Dnipropetrovsk.

Detailed net migration figures by region (in thousands of persons)

Region

Net migration according to Vodafone

Net migration according to the Register

Difference

the city of Kyiv

+57.8

-5.4

-63.2

Kyiv Region

+34.5

+28.9

-5.6

Lviv Region

+31.7

+1.3

-30.4

Zakarpattia Region

+27.4

+4.8

-22.7

Odesa Region

+20.1

+5.8

-14.3

Chernivtsi Region

+16.6

+2.6

-14

Ivano-Frankivsk Region

+15.2

+8.3

-7

Poltava Region

+10.2

+3.8

-6.4

Kirovohrad Region

+8.8

+0.3

-8.5

Rivne Region

+6.3

+2.5

-3.8

Volyn Region

-2.8

+3.3

+6

Cherkasy Region

-3.1

+4.7

+7.8

Zhytomyr Region

-6.6

+1.1

+7.7

Ternopil Region

-8

+3.9

+11.9

Sumy Region

-9.8

-1.9

+7.9

Chernihiv Region

-10.1

+0.1

+10.2

Khmelnytskyi Region

-11.4

+5.2

+16.7

Dnipropetrovsk Region

-12.4

+2.3

+14.7

Vinnytsia Region

-12.7

-0.2

+12.4

Mykolaiv Region

-22.1

-2

+20.1

Zaporizhzhia Region

-27.5

-8.3

+19.3

Kharkiv Region

-102.2

-8.9

+93.3

This means that official figures overestimate population growth compared to data from the mobile operator Vodafone in the Kharkiv, Mykolaiv and Zaporizhzhia Regions, whilst Vodafone “records” significantly higher growth than the Register in the city of Kyiv and the regions of Lviv and Zakarpattia. In other words, the Register simultaneously “records” voters at abandoned addresses in the eastern regions and fails to “notice” their arrival in the western regions and the capital city.

Assessment of discrepancies between data at different levels

Inter-regional breakdown

The Register recorded 0.5 million official changes of electoral address between regions. However, according to Vodafone, actual (extrapolated) inter-regional mobility stands at 1.2 million people. The official coverage rate for inter-regional migration is ≈ 42% (0.5 million ÷ 1.2 million). In other words, around 58% (approximately 0.7 million people) actually live in a different region than that stated in their electoral address.

Since the figure of 1.2 million relates only to movements with a specified destination (and 83% of all migrations recorded by the operator have an unspecified destination), the actual official coverage rate is even lower than 42%.

Outward migration snapshot

Migration abroad for the period 2022–2025 may amount to approx. 5.2 million people. However, the Register data reflects migration abroad via the overseas constituency, to which only 45,000 people have arrived. The official coverage rate for the outward migration is ≈ 1% (0.045 million ÷ 5.2 million). In other words, around 99% of voters reside abroad whilst retaining their electoral address at their pre-war place of residence in Ukraine.

Intra-regional migration

At this level, the figures are reversed: the Register records 1.4 million intra-regional changes of address, compared with 0.64 million according to Vodafone. This does not mean, however, that the official data are more comprehensive. Rather, they are a direct reflection of the different nature of the metrics described in the methodology:

Vodafone counts only those intra-regional moves that crossed a district boundary; a change of address within a single settlement or the same district is not recorded;

The Register, on the other hand, records any change of electoral address, including moves within a single settlement.

Consequently, the significant increase in figures in the official data at this level is due to the broader scope of the metric itself, rather than greater accuracy in recording actual movements.

Possible reasons for discrepancies

The discrepancies identified cannot be attributed to an error in the register; they are due to systemic factors:

  • Voluntary nature of changing one’s electoral address. Updating a record in the Register is a voluntary action on the part of the citizen. Internally displaced persons are not legally required to update their electoral address and use a separate mechanism for temporarily changing their polling station, which is not reflected in the Register’s figures.
  • A consular procedure barrier for migrants abroad. Registration in an overseas constituency requires an application to a Ukrainian diplomatic mission abroad. The majority of the 5.2 million IDPs do not use this procedure.
  • Legal inertia regarding the occupied territories. Although voters have physically left Crimea, Sevastopol, the regions of Kherson, Luhansk and Donetsk, their “old” electoral addresses are still recorded in the Register. The trend in the data for these regions, whilst negative, does not reflect the true state of affairs.
  • The nature of the data collection differs. The Register primarily records a legal reality, which often occurs with a significant time delay, whereas mobile operators’ data records actual locations. The two sources measure different things, “where a citizen is registered” vs. “where a citizen physically resides”.

Geographical patterns of discrepancies

The Register underestimates the influx of people to western Ukraine and the capital city. In the receiving regions, the actual increase significantly exceeds the official figure:

  • Kyiv: Vodafone records an increase of +58,000, whilst the Register, conversely, records an outflow of -5,400; the difference amounts to approximately 63,200 people in opposite directions;
  • Lviv Region: Vodafone data shows an increase of +32,000 compared to +1,300 according to the Register; the difference is 30,000;
  • Zakarpattia Region: According to Vodafone, there was an increase of +27,400 compared to +4,800 according to the Register; the difference is 22,700.

The above regions have hosted a significant number of IDPs who actually live there but have not re-registered their electoral address, regarding their stay as temporary.

At the same time, the Register underestimates the outflow of the population from eastern and southern parts of Ukraine. The actual catastrophic exodus from the frontline regions is barely reflected in the official figures — even though, methodologically speaking, monitoring of subscriber movements in these regions was significantly limited:

  • Kharkiv Region: Vodafone’s figures show a net loss of 102,200, whilst the Register’s figures show a net loss of just 8,900; the difference amounts to 93,300 people (the largest figure);
  • Zaporizhzhia Region: -27,500 in actual figures vs. -8,300 according to the Register; the difference is 19,300;
  • Mykolaiv Region: -22,100 vs. -2,000; a difference of 20,100.

Thus, the population has effectively left the combat zone or areas adjacent to it, but has retained their registered voting address in these regions. The Register still records voters at addresses from which the population has already left.

The discrepancy between the Register and the actual distribution of the population has direct practical implications:

  • Distortions in electoral geography. The discrepancy between the number of registered voters and their actual presence alters the “weight” of the vote and complicates the delineation of electoral constituencies. This poses critical risks to the functioning of a proportional electoral system with open lists, as the size of the electoral quota and the distribution of seats depend directly on the actual number of voters in a specific region.
  • Risks in planning electoral infrastructure. Polling stations at the place of legal registration will be formally overstretched by voters who are not actually present, and understaffed where people actually live.
  • The challenge of organising post-war elections. Without a large-scale update of the electoral register, active voter registration and active promotion of the mechanism for temporarily changing one’s polling station, organising voting for millions of displaced voters is logistically unfeasible.
  • Distortions in socio-economic planning. The distribution of educational, medical and social infrastructure relies on official data – the very same sources from which the Register is largely derived. Due to the critical discrepancy between registration and actual place of residence, state provision in these areas will not meet the real demand on the ground.

Conclusions

There have been 1.2 million inter-regional and 640,000 intra-regional moves, as well as 5.2 million departures abroad, peaking in 2022 (46.2% of cases). Outward migration subsequently stabilised at 0.86–0.99 million people, meaning it became “chronic” rather than ceasing. The figures given represent the lower limit of movements.

The discrepancy between the Register and actual settlement is systemic and critical. The Register reflects less than half of inter-regional migration and around 1% of migration abroad. The registered address of millions of voters does not correspond to their actual place of residence.

In western parts of Ukraine and Kyiv, the Register underestimates arrivals (a difference of 63,000 and a negative net migration balance for Kyiv) whilst in the east and south it underestimates departures (a difference of 93,000 in the Kharkiv Region). The consistency of the net migration rate between Vodafone and the Register is only 64% (14 out of 22 regions).

At the intra-regional level, the imbalance runs in the opposite direction. The Register records more moves (1.4 million compared to 0.64 million according to Vodafone); however, this is not a more accurate reflection of reality, but rather a reflection of the broader scope of the metric: the Register records any change of address, including moves within the same settlement, whereas Vodafone records only movements outside the district.

The reasons for the discrepancies are systemic: the voluntary nature of changing one’s electoral address for IDPs, consular barriers for migrants abroad, legal inertia regarding the occupied territories, as well as the differences in the nature of data recording. None of these factors can be attributed to a technical error in the Register; these are its structural characteristics.

Recommendations

Incorporating data on crossing the state border into the Register data. Given the unprecedented scale of emigration, the Register must be synchronised with the databases of the State Border Guard Service of Ukraine. A legal mechanism needs to be devised to temporarily flag voters who have been abroad for longer than a specified period (e.g. over one year), to ensure the accurate calculation of the number of voters within the country without depriving them of their right to vote.

Regulating the status of electoral addresses in the temporarily occupied territories and in combat zones. A legal framework must be developed to manage the millions of electoral addresses in the territories occupied since 2014 and 2022. Instead of retaining nominal (de facto inactive) addresses, it would be advisable to introduce a procedure for their temporary “freezing”, whilst encouraging voters who have moved to government-controlled territory to register at their new actual place of residence.

Encouraging the proactive updating of electoral addresses. The state must change its approach to voter registration, making it as accessible as possible. An extensive information campaign must be launched, and tools for changing electoral addresses must be integrated into the DIIA ecosystem to encourage migrants and IDPs to update their details well in advance of the start of the electoral process.

Conducting a nationwide study of mobility involving all mobile operators. The State has every opportunity to initiate its own more comprehensive study of subscriber movement using aggregated anonymised data from the three largest mobile operators (Vodafone, Kyivstar and Lifecell), which will provide accurate absolute figures rather than extrapolation errors. The resulting dataset must be cross-referenced with key state databases recording citizens’ actual whereabouts (registers of the State Border Guard Service, IDPs, the Social Security Administration, subsidy recipients and insured persons). Such a step, implemented on a sound legal basis and in compliance with data protection laws, will provide an accurate official framework for verifying and updating the electoral register ahead of the post-war elections.

Conducting a comprehensive audit of the Register ahead of the post-war elections. As the current Register has lost its relevance as a reflection of the “actual electoral geography”, the first step in preparing for the post-war electoral cycle should be a nationwide update of the Register. All available sources and cross-register verification should be used to update the data. Citizens’ actual places of residence can be confirmed through other state databases that record a person’s actual whereabouts: databases of IDPs and recipients of housing subsidies, the Electronic Health System (medical declarations), the register of insured persons (place of employment/tax payments), etc.

The project is being implemented by the OPORA Civil Network with financial support from the German Federal Foreign Office.