Population projection (experimental statistics)
1. Contact
Responsible agency
Unit
Contact person
Position
Email (agency)
Phone
2. Statistical presentation
Data description
Experimental statistics showing population projections for the administrative territories of Latvia, broken down by sex and age group. The projections are based on demographic trends in fertility, mortality and migration recorded for the period 2017–2024 and reflect expected population developments assuming these trends continue.
Experimental statistics are produced by using new data sources and methods in making attempts to expand the range of statistics or the level of detail thereof based on the needs of data users.
It should be noted that the methods used in experimental statistics are not constant, approbated or internationally harmonized and can be changed to improve data quality. CSB publishes experimental statistics to get user feedback, evaluate analytical potential of the data and relevance thereof to the actual reality and data user needs. The CSB believes that data users may find experimental data series valuable, and opinion of the data users may serve as a basis for making a decision to include these statistics in the Official Statistics Programme. By publishing experimental statistics, the CSB provides data users with new sources of information that may be used for decision making.
Calculations are performed separately for each sex and age cohort using the cohort-component method, by sequentially applying fertility, mortality and migration rates. Data are compiled for the territorial units in force after the administrative territorial reform implemented in 2021. These units include: State cities and State city neighbourhoods, municipality towns and municipalities, rural territories (pagasti), and densely populated areas (DPA).
Within a given year, the place of residence is the location with the longest overlap with the period of employment. Accordingly, a person may have more than one place of residence within a year. If a person has no place of employment, the place of residence with the longest duration during the year is used.
Age is calculated using the completed calendar year method (age in completed calendar year). Age = reference year − year of birth, regardless of the month and day of birth. This corresponds to the age the person reaches by the end of the year.

Project "Geospatial statistics – functional areas and demographic projections in territories", Grant Agreement No. 101195934.
Statistical concepts and definitions
Statistical unit
Usual resident of Latvia.
Statistical population
All usual residents of Latvia, i.e., persons living there or having an intention to live there for at least 12 months.
3. Institutional mandate
Legal acts and other agreements
N/A
4. Accessibility and clarity
On-line database
5. Comparability
Comparability - geographical
N/A
Length of comparable time series
Projections are not directly comparable over time, as coefficients are re-estimated with each new data period.
6. Coherence
Coherence- cross domain
N/A
7. Statistical processing (data source etc.)
Source data
- Register of Natural Persons (until 28 June 2021 Population Register) managed by the Office of Citizenship and Migration Affairs
- Boundaries of neighbourhoods and densely populated areas available on the Latvia's Open Data Portal
Data collection
Data are collected from administrative registers.
Data compilation
Calculation principles
Projections are based on the cohort-component method, a demographic projection approach in which the population is divided into 202 cohorts (2 sexes × 101 age groups: 0–100). Three components are applied to each cohort:
- fertility – births estimated from women of reproductive age and recorded fertility rates;
- mortality – deaths estimated using recorded mortality rates;
- migration – immigration and emigration estimated for each sex and age group.
Calculations are carried out sequentially: each year, individuals are reassigned to the next age group, and the corresponding rates are applied to each cohort. The population is updated according to the following formula:
P(t+1) = P(t) + Births − Deaths + Immigrants − Emigrants.
Coefficients are calculated from historical data (2017–2024) and assumed constant during the projection period.
Age groups are defined based on the average age of the population during the reference period, rather than age as on 1 January.
Fertility rates
The age-specific fertility rate (ASFR) expresses the number of live births per woman of a given age during the reference period:
\(ASFR_a = \frac{B_a}{\bar{P}_{F,a}}\)
where Ba is the number of live births to mothers aged a, and P̄F,a is the average number of women aged a. The rate is calculated for maternal ages 15 to 49.
The total fertility rate (TFR) represents the average number of children a woman would have during her lifetime if age-specific fertility rates remained at the level observed in the reference period:
\(TFR = \sum_{a=15}^{49} ASFR_a\)
The sex distribution of births is determined using the recorded proportion of male births (rmale = 0,5142). Infant mortality is applied to newborns.
Mortality rates
The age-specific mortality rate (ASMR) expresses the number of deaths for the given sex and age group relative to the average population of that group during the reference period:
\(ASMR_{g,a} = \frac{D_{g,a}}{\bar{P}_{g,a}}\)
where Dg,a is the number of deaths for sex g at age a, and P̄g,a is the average population of that group. The rate is calculated separately for each single year of age and sex.
Migration rates
The immigration rate expresses the proportion of persons who entered the territory during the reference period and did not reside in it in the previous period (excluding live births):
\(IMM_{g,a} = \frac{I_{g,a}}{\bar{P}_{g,a}}\)
where Dg,a is the number of immigrants by sex g and age a, and P̄g,a is the average population of that group.
The emigration rate expresses the proportion of persons who resided in the territory in the previous period but no longer reside in it in the reference period (excluding deaths):
\(EMI_{g,a} = \frac{E_{g,a}}{\bar{P}_{g,a}}\)
where Dg,a is the number of immigrants by sex g and age a, and P̄g,a is the average population of that group.
Migration rates are calculated as the average of all available reference periods.
Projection variants
Population projections are produced for a baseline variant, while alternative projection variants can also be defined by adjusting a set of assumptions about fertility, mortality, immigration and emigration. These adjustments (scenarios) are implemented through multipliers (coefficients) applied to the corresponding demographic rates.
Each multiplier ranges from 0.9 to 1.1 (±10%). A value of 1.0 represents the baseline variant, where current trends continue unchanged. Different combinations of multipliers define alternative variants (e.g. pessimistic or optimistic).
For example, a value of 1.1 for fertility corresponds to a 10% increase in fertility rates, while 0.9 for emigration corresponds to a 10% decrease in emigration rates.
Adjustments
In small areas, observed demographic rates may be statistically unstable, as a low number of events can result in large fluctuations. To ensure the robustness of projections, the following adjustments are applied:
Replacement of fertility rates: if no births are observed in a given maternal age group within a territory and the calculated age-specific fertility rate is zero, it is replaced with the corresponding national-level fertility rate. This prevents zero fertility being projected for the entire projection period due to random variation.
Replacement and bounding of mortality rates: if no deaths are observed in a given sex and age group and the age-specific mortality rate is zero, it is replaced with the corresponding national-level mortality rate. In addition, a constraint is applied: ASMRg,a ≤ 1,0, to prevent the rate from exceeding 100% in small groups with multiple deaths.
Migration rates are not adjusted using national-level values. Neither replacement nor bounding is applied, as migration intensity may differ substantially across territories and these differences are preserved.
Non-negativity constraint: during the calculations, cohort population counts are constrained to be non-negative (the minimum value is zero).
Reference periods for coefficient estimation
The most recent available periods are used to estimate coefficients (configurable; currently 2017–2024).
Exclusion of COVID-19 impact
The COVID-19 pandemic resulted in unusually high mortality in 2021–2022, which does not reflect long-term demographic trends. These years are excluded from the calculation of age-specific mortality rates.
Exclusion settings:
| Parameter | Value |
| Excluded years | 2021, 2022 |
| Affected indicator | age-specific mortality rate only |
| Unaffected indicators | age-specific fertility rate, migration |
If the exclusion reduces the number of available periods to fewer than one year, all periods are used automatically.
Actual age-specific mortality rate calculation periods: 2017, 2018, 2019, 2020, 2023, 2024.
Assignment of place of residence for newborns
Newborns are assigned to the territory of the mother’s declared place of residence rather than the child’s place of registration. This approach more accurately reflects the actual population distribution, as infants typically reside with their mother, and the mother’s location determines demand for childcare and related services.
The method is consistent with UN and Eurostat recommendations and covers approximately 99.9% of birth cases. In cases where the mother’s place of residence is unknown, the child’s place of registration is used.
Maternal age range used for fertility rate calculations: 15–49 years.
Expansion of small territories
For small territories (population < 30 000), demographic rates may be statistically unstable due to a low number of events. To improve the reliability of coefficient estimation, an automatic territory expansion procedure is applied.
The procedure is as follows:
- The population centroid of the territory is determined.
- If the population is below 30 000, the radius is incrementally increased.
- The radius is expanded until the covered population reaches at least 30 000 or a maximum radius of 100 km is reached.
- If the required population threshold cannot be reached within the maximum radius, national-level rates are used.
The expanded zone is used only for the calculation of coefficients (used with ASFR, ASMR, migration rates).
The base population always consists of the residents of the respective territory.
The radius is determined as:
\( r = \min\!\left( r^* :\ \displaystyle\sum_{i \,\in\, B(c,\, r^*)} P_i \geq 30\,000,\quad r^* \leq 100\,000 \right) \)
where B(c, r*) is the area (circle) with centre c and radius r* (in metres).
Replacement with national-level values
If data are insufficient to estimate rates for a territory, even after expansion, the corresponding rates are replaced with national-level values:
• national age-specific fertility rates by age of mother;
• national age-specific mortality rates by sex and age.
The sex distribution of births is determined using the national proportion of male births (rmale = 0,5142).