The Continuous Population Statistics (CPS) is an operation of the Instituto Nacional de Estadística (INE), the national statistics institute, that provides quarterly figures for:

• Population residing in Spain
• Population residing in family dwellings
• Households
• Quarterly estimates of external migration

The CPS is part of the new statistical production system based on administrative records, which allows for the generation of demographic information with greater frequency, timeliness and consistency compared to traditional methods based on surveys or decennial censuses.

Thanks to this approach, the CPS makes it possible to:

• Measure demographic dynamics more immediately, offering an updated and continuous view of population changes.
• Observe these changes almost in real time, including modifications in population and household structure, as well as recent migratory movements.
• Build a complete historical series of the resident population in Spain, linking the different censuses and providing provisional estimates based on the last survey published.

Overall, the CPS is a key instrument for the quarterly monitoring of Spanish demographic developments within a modern statistical system, based on the efficient exploitation of administrative records.

Each type of series has a construction method:

• Population residing in Spain:
  • The final data are composed of population censuses and intercensal estimates between two consecutive censuses.
  • The provisional data are constructed by applying the changes observed in the variations that occurred in the population register to the latest published population census. Births and deaths from the Civil Registry may be recorded with some delay in the census database. To correct for this effect, provisional birth and death data, collected respectively in the statistics, are used: Estimated number of weekly deaths and Monthly birth estimate.
• Population residing in family dwellings and households: This information is obtained from the housing census. For dates after that census, the proportions observed in the last published housing census are applied to the resident population.
• External migrations: a direct exploitation of the population variations associated with migratory patterns. 

• Population residing in Spain: although the CPS began as a new operation starting from the 2021 Population and Housing Census, it also collects all previous resident population information: that corresponding to previous censuses, as well as their intercensal estimates. Therefore, the comparable resident population series begins at two different times, depending on the detail provided by previous censuses: 1971 or 2002.

  From May 2026, the resident population of the main municipalities and provincial capitals as of January 1 of each year is also provided. Since no intercensal estimates were made prior to 2021 at the municipal level, this series begins in 2021.

• Population residing in family dwelling: since 2021, because no previous intercensal estimates were made.
• Households: since 2021, because no previous intercensal estimates were made.
• Quarterly estimates of external migration: since 2023, which is when they started to be assembled.

The CPS provides information with the following level of territorial detail:

• Resident population: province and island. Since May 2026, it also provides the population as of 1 January for the main municipalities and provincial capitals.

  This approach lets consistent coverage be combined at the provincial level with higher resolution in those municipalities that concentrate a significant part of the population or that have a relevant administrative role.

• Population in family dwellings, households and external migrations: province.

The CPS provides detailed quarterly information on various demographic and migratory aspects of the population residing in Spain. Specifically, it offers:

• Total resident population, broken down by:
  • Sex
  • Age
  • Nationality (Spanish / foreign for provisional data)
  • Country of birth (Spain / foreign for provisional data)
• Population residing in family dwellings, with the same level of disaggregation as the total population, which provides knowledge of the structure and demographic characteristics of those who live in this type of dwellings.
• Total number of households, classified according to household size, facilitating the analysis of the composition and evolution of households in Spain.
• External migrations, with quarterly information for the main countries of nationality and birth, which allows for the evaluation of the migratory situation and its impact on population changes.

The CPS is published on a quarterly basis, approximately one and a half months after the end of the reference quarter (in February the data for the last quarter of the previous year, in May the first quarter of the year, in August the second and in November the third). Since the last census was published, the figures are provisional, being revised in each publication as new administrative records are integrated and existing ones are cleansed.

For example, in May of year T, a new reference date is published, 1 April of year T, and the entire provisional series is revised at that time, from 1 April of year T-1 to 1 January of year T.

This review process ensures that the entire series maintains internal consistency and that the quarterly figures are aligned with the Annual Population Census, which constitutes the highest quality reference point within the statistical production system based on administrative records.

Population data for the main municipalities and provincial capitals are excluded from this process, and are only published in the May edition of the CPS, referring to 1 January of that year.

Different review policies are applied, depending on the nature of each statistical series:

• Resident population:

  These series are initially published as provisional and are fully revised each quarter (except for those of municipalities, which are only published as provisional once), incorporating the new available census and administrative information.

  They become definitive when the Annual Population Census for the corresponding year is published.

• Population residing in family dwellings and households:

  It is revised when:

  • The resident population (whether provisional or permanent) is updated, as this is its starting point, then applying the proportion of family dwellings or households observed in the last housing census.
  • A new housing census is published. The last one was the 2021 Population and Housing Census, and the next Housing Census will take 2026 as the reference year.

  Similarly, a complete review of the entire provisional series is carried out each quarter.

• External migration

  The quarterly migration series provided by the CPS is not reviewed.Its purpose is to capture current trends, so it remains a fixed estimate once published.

The resident population is the total population that usually resides in a place.

In turn, each person can reside in a family dwelling, or in a collective establishment. In other words, a family dwelling is one intended to be inhabited by one or more people who do not constitute a group.

So, the (resident) population in family dwellings is made up of all the residents who live in family dwellings, and are distributed in households. Households in the CPS are defined according to the criterion of residence, that is, a household is the human group formed by one or more people who usually reside in a family dwelling.

Note that a family dwelling may or may not be the habitual residence of a group of people, so a household is always a family dwelling, but a family dwelling does not necessarily constitute a household.

They are not the same households, for several reasons which are detailed below.

Surveys calibrate their households to the results of the CPS at each moment, so at the time of calibration, the first estimate provided by the CPS, one of the subsequent revisions, or even the definitive value may be available. However, surveys don't usually revise the series.

According to European regulations, all European surveys targeting the population (residents in family dwellings) must use the budget criteria to build homes, so that the home is determined not only by common residence but also by sharing overhead costs and budget.

However, as explained in a previous question, the CPS uses the residency criteria so that a household consists of all the people living together in the same family dwelling.

With the sources used to construct the CPS mainly the population register and the censuses, it would not be possible to obtain the information necessary to construct households according to the budget concept, which is only available in the surveys.

For this reason, a specific estimate is made for the surveys to transform the residence criterion provided by the CPS into a budget concept.

In practice, the number of households according to the residence criterion will be lower for the same population residing in family dwellings, and they will have a greater number of members, than those that follow the budget criterion.

It is worth noting that the largest of the surveys aimed at households, the Active Population Survey (APS), introduced the concept of household according to budget in its 2021 series.

In addition, it has a very demanding publication schedule, and requires an estimate of households one month prior to publication by the CPS. To do this, the INE makes an ad hoc advance estimate for the APS, which never revises its series.

Furthermore, the households resulting from the APS refer to the middle of the quarter, while those from the CPS refer to the end of the quarter. 

The CPS offers a quick measurement of external migration that occurred in the previous quarter, based on the population register variations observed in that period that are associated with migratory movements. Its main purpose is to quickly detect recent trends -for example, the emergence of a new country of relevant nationality among migratory flows- and to provide a near-immediate view of the current migratory situation.

However, to accurately analyse the actual level of migration of the resident population and to have more complete and disaggregated information, it is necessary to use the Statistics on Migration and Changes of Residence (SMCR).

The SMCR is the reference operation for comprehensively measuring internal and external migrations, but it is available with a greater time lag, as it requires additional processes of integration, verification and cleansing of the information from the population and the Annual Population Census.

This is why the migration information from the CPS should not be interpreted as an advance of the SMCR.

Both operations are complementary:

• The CPS provides speed and the ability to detect recent changes.
• The SMCR provides accuracy, detail, and statistical stability.

The CPS and the Annual Population Census are part of the same statistical production system based on administrative records, and are closely linked.

The CPS offers a complete historical series of the resident population in Spain, linking the different censuses and also providing provisional estimates based on the last survey published. Its main contribution is that it allows us to know the evolution of the population on a quarterly basis, even before the next census is published.

For its part, the Annual Population Census provides the highest level of territorial detail (down to the census section) and sociodemographic characteristics available each year, and constitutes the highest quality official reference of the system. However, the annual census does not include information on the population residing in family dwellings or on the number of households, aspects that the CPS does provide on a quarterly basis.

Furthermore, the Annual Population Census plays a fundamental role in the construction of the provisional series, as it is taken as a starting point to incorporate the changes observed in the census records thus far.

In short:

• The CPS provides immediacy and quarterly monitoring, and provides the complete series of resident population.
• The annual Census provides accuracy, depth, and detail.
• Both operations are complementary and consistent with each other within the statistical system based on administrative records.

When the municipal figures from the CPS are definitive, they do coincide with those of the Annual Population Census. The differences between census figures and population register figures are explained in detail in this latter operation.

When these figures are provisional, they differ even more, since the starting point is the most recently published population census, and the general method of estimating provisional data is added to it.

Traditionally, after the publication of each ten-year census, a statistical population series would start. Specifically, starting with the 2011 Population and Housing Censuses, and beginning with 1 January 2012, the Population Figures series was published, estimated by means of accounting for demographic events (births, deaths, internal and external migrations and acquisitions of Spanish nationality by the foreign population).

However, when a new census is published every ten years, the population data series for the entire prior decade is corrected up to the previous census (traditionally called intercensal estimates).

With the availability of annual censuses, this correction is no longer necessary. Furthermore, this initiated a new dynamic in population estimation processes, so it was considered appropriate to create a new operation (the CPS), different from the old PFs, to incorporate the 2021 census, the 2011-2021 intercensal estimates, and continue the process from there.

In summary, the main differences between CPS and PF are:

• Until 1 January 2012, the resident population series of the PF coincided with that of the CPS. With the publication of the 2021 Population and Housing Census, the PFs were no longer updated, nor was the 2011-2021 intercensal estimate. This updated data had already been collected in the CPS, which then continued with current estimates:
• The construction of the PFs was based on flows or forward flows (application of the demographic phenomena of a period to a starting population to obtain a population at the end of the period). However, the new methodological approach to CPS is based on stocks: the best possible populations are obtained at a given date (censuses), and demographic flows between two censuses are adjusted at a later date.
• There was no annual baseline census at the PFs to correct them; There was only one revision every 10 years, when we had the new census (the traditionally called intercensal estimates). The CPS has annual censuses, so that ten-year review is not necessary.
• Reference dates: in the PFs there were only two per year (1 January and 1 July), while in the CPS there have four reference dates (1 January, 1 April, 1 July and 1 October).
• Provisional data: in the PFs the first provisional data were given about 5 or 6 months after the reference date, while in the CPS they are about 45 days later. However, the level of detail of the provisional data in the PFs was the same as that of the final data, whereas in the CPS they have a lower level of detail.
• The PFs only provided the resident population series, whereas the CPS also provides the resident population of family dwellings, and the number of households, as well as initial external migrations that can explain the population change in each quarter.

The Continuous Household Survey investigated the basic social and demographic variables, both for households and the people in them and the dwellings they inhabit.

The CPS only provides the number of households by province and size, and not other household characteristics that the CHS investigated.

The CHS data series spans from 2013 to 2020, and was calibrated to the Population Figures (which also provided an estimate of households for the calibration of the surveys, although it was not published), so its series is not up to date after the 2021 Population and Housing Census.

As a result, the household series from the CPS and the CHS are not comparable.