455. Geospatial and statistical data governance is crucial to avoid duplication of work and to enable sharing of data. Geospatial data plays a key role in revealing insights, patterns and trends, and thereby supports decision-making.

456. For the use of geographic information systems (GIS) in censuses, the following benefits and costs need to be considered:

(a) Benefits:

i. Closer linkage between maps for enumerators and map-based products for users;

ii. Enriched dissemination of census data as they can be visualized with reference to geographic areas for easy understanding by users;

iii. The cost of intercensal updating of the base map is lower with a digital base map, enabling, among other things, the construction and updating of sampling frames;

iv. Producing duplicate maps may be less expensive with a GIS solution;

v. GIS has increased ability to undertake quality assurance of geographic boundaries;

vi. GIS allows NSOs to have a greater ability to perform spatial queries and advanced analyses;

vii. The space needed to store input maps for digital purposes is much reduced.

(b) Costs:

i. GIS requires additional technical expertise;

ii. GIS requires a higher level of computing infrastructure;

iii. A clerical census system can proceed on the basis of basic maps. However, use of GIS in this task requires that a digital map base exists. If it is necessary to create the digital map base, which requires significant lead times as well as significant funding. In both cases, more experienced technical staff are required;

iv. In most cases, the preparation of maps or GIS are not the core business of a statistical office.

457. Census information at accurate and standard geographic levels is essential to facilitate comparative analysis and achieve better quality of geospatial statistical data production. In the context of a census, it is crucial to have georeferenced data at the level of x, y coordinates.

458. The location of a given census unit should be described using coordinates (in the coordinate system commonly used in the country), especially taking into account the register identifiers of territorial units and address features (including the identifiers of administrative division units, locality identifiers and street identifiers).

459. For the purpose of conducting enumeration and disseminating results, the address point coordinates should be determined with an accuracy of several metres.

460. With the aim of facilitating the management of spatial data concerning the location of census units, the coordinate system adopted should be homogenous for the whole country or territory.

461. A geographic information system (GIS) can be seen as a system of hardware, software and procedures designed to support the capture, management, processing, analysis, modelling and display of spatially-referenced data.

462. Careful planning should be undertaken to plan for a complete GIS that can solve complex census operational planning issues, providing solutions for managing operations, identifying, and resolving enumeration coverage problems, monitoring field operations, enforcing quality controls, producing detailed georeferenced inventories, and supporting a variety of dissemination platforms, products and formats.

463. GIS technology should be considered only at a level appropriate to the skills and resources available and should constitute an integral part of the overall work of the organization. Cooperative arrangements with a national mapping agency or other agencies should be pursued, particularly regarding the acquisition and maintenance of base map data, which usually is not the responsibility of the statistical office. Therefore, as many administrative sources as possible should be reviewed for presence of spatially-referenced data and tested for the possibility of using it for statistical purposes. The United Nations Integrated Geospatial Information Framework (UN-IGIF)³⁷ provides a basis and guide for developing, integrating, strengthening and maximizing geospatial information management and related resources supporting NSOs in seeing themselves in the national geospatial data production ecosystem.

464. Given the great potential of geospatial information for censuses, statistical offices should pursue the development or implementation of GIS and integration of geospatial information whenever possible, including for intercensal surveys. The integration of statistical and geospatial information must be an objective of NSOs, understanding themselves as important institutions that produce and consume geospatial information.

465. In recent years, many countries have adopted GIS to facilitate census mapping in the production of enumeration maps and dissemination products. As their cost is decreasing, geospatial technologies and inputs are being made available free of charge or openly, and the basic technology is now well established. This trend can be expected to continue and expand in future census rounds. It is likely that the census could help to increase the capacity of the statistical office (or the country as a whole). The adoption of GIS should therefore be seen as an important strategic decision, with impacts that go beyond the operation of the census.

466. The scope of geospatial and statistical data integration is very complex. The challenge is to understand how to achieve this integration in the most effective and consistent way. Developing a coherent and systematic approach for linking statistical and geospatial data requires considerable commitment and time.

467. The best way to achieve a consistent integration is having a common method for enabling statistical and administrative data to be geospatially referenced, preferably in connection with the Global Statistical Geospatial Framework (GSGF),³⁸ which enables comparisons within and between countries. The GSGF framework consists of five principles that are considered essential to integrate statistical and geospatial information. Moreover, GSGF is a high-level framework which facilitates consistent production and integration approaches for geo-statistical information. It is a generic framework that permits the application of the principles to local circumstances.

468. The creation of a geocoding (see section ‎8.7) infrastructure for statistics and its integration into the statistical production process does not demand a complete redesign of enterprise architectures and statistical production processes. Small and stepwise improvements are possible. However, integration of statistical and geospatial information is a cornerstone in the modernization of official statistics.

469. Census information at accurate and standard geographic levels is essential to facilitate comparative analysis and achieve better quality of geospatial statistical data production. In the context of a census, first of all it is crucial to have georeferenced data at the level of x, y coordinates. The introduction of x, y coordinates and address points in census data enables a shift from area assignment (census districts) to point assignment. The change of the assignment mode allows for more flexible grouping and presenting data collected in statistical units smaller than the commune, i.e. in statistical regions, census enumeration areas or even in very small areas, such as a grid with a cell size of one square kilometre. It also facilitates the creation of spatially-oriented microdata that would enable geo-statistical analysis in a user-defined area.

470. Wherever possible, data should be collected with reference to an address point; results can then be disseminated using any desired spatial divisions.

471. The address point approach refers to the strategy of geocoding census data using addresses to facilitate geocoding. Geocoding address data is usually done through specialized software or via an address register. In certain cases, manual geocoding of certain addresses that have not been geocoded otherwise must be added.

472. Geocoding of address points offers the following advantages:

(a) Addresses are a common part of data that is used or collected for census purposes. Geocoding this data can therefore be done in existing workflows with minor adjustments. Connecting this address data to geographical coordinates broadens the possibilities for analysis. Address points offer a high level of detail that presents analytical benefits. In addition, using the address point approach is robust against postal and administrative restructuring, and is therefore more suited to comparing data from different census rounds;

(b) The availability of a precise coordinate for the location of each household can tie each individual response to a precise location where it was collected, which offers great flexibility for disseminating results by geographic units.

473. Mapping has been an integral part of censuses for a long time and continues to be widely used irrespective of the type of census. Over the years, census maps have played a critical role in all processes from preparation to dissemination of census results.

474. The census-mapping programme should be developed at a very early stage of census planning, considering the conditions and available resources. Statistical offices should evaluate available mapping options by considering the following factors:

(a) Available geographic resources;

(b) Requirements for new technologies and approaches;

(c) Available funds and allocated time frame;

(d) Staff capacity needed for new approaches; and

(e) Requirements for developing an ongoing geographic system.

475. Evaluation of these factors will determine the best mix of technology and other approaches for each individual case.

476. Today maps play an important role, especially in the dissemination of census results, as they allow the main characteristics of the population and housing units in the different territories to be represented. Users are increasingly interested in census statistics at a very detailed territorial level.

477. Another important aspect is to strengthen collaboration between NSOs and mapping agencies for the development of a common GIS and a geospatial framework. The development in these areas would produce numerous benefits, such as working on shared geographic areas, standardizing spatial concepts and codes, facilitating data integration and improving the dissemination of census data at finer spatial levels.

478. Integration of statistical and geospatial information consists of the use of geospatial information for the production and dissemination of statistics. Integration can take place at any stage of the statistical production process, as described by the Generic Statistical Business Process Model (GSBPM)³⁹ and in more detail in the geospatial view of it (GeoGSPBM).⁴⁰ The integration includes geocoding of statistics, spatial analysis and statistical maps. As part of the integration process the following steps may occur:

(a) Geocoding statistical information at unit-record level;

(b) Processing statistical information using spatial analysis techniques with the purpose of selecting information or deriving new information with a focus on spatial characteristics, e.g. buffering around spatial features;

(c) Supporting a more efficient and flexible statistical production process with geospatial information, e.g. for surveying and sampling, in the case of field operations;

(d) Combining final statistical data with geospatial information on the level of database gives the possibility to conduct geospatial analyses and to present the results on statistical maps;

(e) Improving the quality of existing statistical products by adopting spatial models, e.g. producing information on commuting patterns by calculating journey times based on detailed transport networks.

479. All statistical phenomena that can be associated to a location are in principle relevant for the integration of statistical and geospatial information. Location in this context means the location of the individual observation at unit record level. In most cases the location will be a point with coordinates or a precise address. However, other spatial reference frameworks such as lines or polygons are relevant as well representing, e.g. road segments or areas with a certain land cover.⁴¹

480. Cooperation and collaboration on integration of census data and geospatial information can be a viable strategy to utilize the existing capabilities of other national actors. The opportunities and constraints of collaborating or cooperating with other actors depend heavily on the local statistical and administrative system. The roles and responsibilities of the involved actors therefore must be adapted to the local systems.

481. Regardless of the structure of the statistical geospatial framework, it is important to record the data only once to reduce errors and enable synergies.

482. Both NSOs and national mapping agencies play an important role in data governance. The integration of statistical and geospatial data has been playing an increasingly important role and offers one of the most promising paths to provide reliable and detailed information including census information about population and housing that can result in new insights.

483. A common set of geographies based on traditional statistical and administrative geographies should be used to display, report and analyse census information.

484. According to the GSGF, a common set of geographies ensures the consistent geospatial aggregation, comparability and dissemination of statistical data. Data are uniformly allocated to smaller administrative segments or statistical units such as mesh blocks that are divided according to political, property, or topological subdivisions.

485. Aggregation and disaggregation methods enhance quality, assessment, consistency, comparability and use of data.

486. The common geographies should be based on authoritative data and also be included in the National Spatial Data Infrastructure (NSDI). The governance and management of common geographies include the agreed scale, reference dates, coding systems, resolution and accuracy of geographies. The framework of geographies also needs to be developed continuously in line with emerging user needs.

487. The concepts for understanding the delineation methodologies used to maintain common dissemination geographies are potentially complex, but it is important that they are understood and used by stakeholders. For this purpose, providing documentation on the delineation methodologies for the dissemination geography and its lifecycle and changes is essential.

488. Geocoding is the process of transforming a description of a location or location information (such as an address, name of a place, or coordinates) to a location on the earth’s surface. In other words, geocoding is a way to ensure “data know where they are”. It is also defined as translation of one form of location into another (ISO 19133:2005).

489. For the purposes of the GSGF, geocoding is generally defined as the process of geospatially enabling statistical unit records so that they can be used in geospatial analysis.

490. More specifically, geocoding is the process of linking unreferenced location information, often in the form of a text string (e.g. an address or address identifier), that is associated with a statistical unit, to a geocode. Alternatively, the geocode can be incorporated directly into the statistical unit record.

491. The condition for geocoding is the availability of high quality physical address, property or building identifiers, or other location descriptions, in order to assign accurate coordinates and/or a small geographic area to each statistical unit.⁴²

492. Geocoding refers to the process of linking geographical coordinates to address information. Geocoding is therefore the prerequisite to processing address information to make it usable in a geographic context. A register of addresses, buildings and dwellings would provide a suitable basis for such geocoding, and it would therefore be important to keep it up-to-date.

493. A high-quality, comprehensive, updated and georeferenced address register of each building and dwelling can give great support in planning and organizing a census. It can play a central role in many fieldwork operations and provide the key to accurate delivery, collection and follow-up of questionnaires.

494. The best way to associate each address with a location in physical space of a map is to specify its coordinates in a proper geographic reference system. With geographic coordinates, addresses can be entered in available maps or into the GIS. If it is not possible to get coordinates it is recommended at least to geocode addresses by finding associated geographic coordinates from other geographic data. For example, as geographic coordinates of an address, the coordinates of the centre point (centroid) of the enumeration areas to which the address belongs could be taken.

495. It is crucial to consider carefully how this linkage is performed and to ensure the geographic accuracy of the resulting data. Typically, buildings are sourced from administrative records in a geographic format. It is advisable to verify that the distance between a building and the coordinate used for its address is consistent with the reality. There may be instances where an address has an estimated coordinate that is distant from the building it represents.

496. To increase acceptance and usability of a register, it is necessary to include a common and stable identifier (ID). This would greatly facilitate data exchange, cooperation, updating geospatial information and the production of related statistics.

497. Digital maps can be used for the planning, preparation and management of the census, and throughout all the stages of conducting the census, including pre-enumeration.

498. With a view to drawing up the census documentation on digital maps, performing geo-statistical analyses and presenting output data on digital maps, a pre-enumeration assumption should be made that appropriate GIS tools compatible with spatial microdata will be used already at the stage of delineation of enumeration areas.

499. It is necessary to design the databases comprising the locations of census units, which should be obtained from the NSO or the mapping agency.

500. It is necessary to verify the correctness of data concerning census unit locations. Such verification may be performed on the basis of data coming from administrative sources ‒ a referential register comprising address points usually kept by the mapping agency.

501. If no such administrative sources are at hand, the accuracy of data concerning census unit locations may be verified at the pre-enumeration stage, during which the enumerators' task is to verify and adjust (where necessary) the spatial location of individual census units, using handheld devices equipped with GPS.

502. Accurate data on census unit locations ensure efficient management of enumerators' work, in terms of assigning census questionnaires, monitoring enumerators' activities and monitoring the progress of the census enumeration at each stage.

503. It is crucial to obtain the locations of census units as x, y coordinates. The location of a given census unit should be specified by means of coordinates and descriptive attributes, especially taking into account the register identifiers of territorial units and address features, including the identifiers of administrative division units, locality identifiers and street identifiers.

504. For the purpose of conducting the enumeration and disseminating results, the address point coordinates should be determined with an accuracy of several metres (no more than seven metres).

505. With the aim of facilitating the management of spatial data concerning the location of census units, the coordinate system adopted should be homogenous for the whole country.

506. Production of digital maps of enumeration areas may be done by using data from authoritative geospatial data and services such as national spatial data infrastructures. The acquisition of census locations makes it possible to design the map services and dedicated applications to display maps by both enumerators and their supervisors. Below is a sample content of the digital map to be used by census supervisors. It can consist of the following layers:

(a) The administrative division boundaries;

(b) The statistical division boundaries;

(c) Cadastral data: cadastral parcel boundaries, building contours (depending on the completeness of acquired data);

(d) Road and street network;

(e) Address points (census unit locations along with x, y coordinates);

(f) Some topography elements such as bodies of water and railway lines

507. An orthophoto map and orthoimages obtained from very high spatial resolution satellite images may serve as the background for these layers.

508. The digital map enables an interactive allocation of workload for field workers using the GIS technology. The visualization of census units against the division into enumeration areas fosters the accurate determination of the workload to be performed in a given enumeration area and its adequate division among enumerators.

509. The location of census units can then be verified during the pre-enumeration round. The pre-enumeration round could be aimed at verifying the existence of buildings and supplementing the list with missing address points. Additionally, during the pre-enumeration round enumerators have a chance to familiarize themselves with the area where they will later perform enumeration.

510. During the pre-enumeration round enumerators could not only confirm, remove or modify building data in respect of the address points assigned, but also inspect the entire census areas entrusted to them.

511. GIS technology can be used to assign work to individual enumerators, as well as to monitor their activities and provide the necessary assistance. The tasks are assigned on the basis of the statistical division of the country. The GIS system can be used to entrust enumerators with the tasks located in the vicinity of their place of residence. However, it is a good practice to assign enumerators to census areas other than those containing or in the immediate vicinity of their direct places of residence.

512. Census supervisors of the teams of enumerators should provide front-line support to enumerators. They should also be equipped with a desktop GIS tool which will allow them to monitor the fieldwork performed by enumerators.

513. Enumerators could use a map module to find their way to the location of the address. Due to the continuous access to the map and the possibility to determine their locations, enumerators will be able to effectively plan their work and visit units located close to one another on the same day.

514. With the aim of ensuring enumerators' safety, the mobile application on hand-held devices should be equipped with an alert function to notify the local leader or supervisor of any emergency situations and to provide support (emergency service, police) through an embedded GPS module.

515. The census progress and completeness could be also monitored on maps. Visualizing the census units on maps makes it possible to monitor the census completeness at any aggregation level (i.e. by census area, statistical region, local level or regional level), and to verify each census unit individually, in terms of the correctness of the conduct of the census.

516. While monitoring the census completeness, supervisors can monitor, on an ongoing basis, the workload of each enumerator and decide on reallocating tasks between individual enumerators, with the aim of ensuring a smooth course of census operations.

517. The GIS module also makes it possible to monitor enumerators' work in real time by displaying their routes and verifying whether the completed questionnaires are sent to the system from the locations where interviews have been conducted.

518. Most countries are able to geocode census information at the level of individual geographic coordinates and subsequently obtain aggregate data at any spatial level and even at the 1km² grid level.

519. Providing grid-level census data is not only useful to better understand small territories, even regardless of administrative boundaries, and to study local dynamics, but it is necessary to enable policy actors to adopt intervention policies based on the specificities of local realities.

520. From the point of view of Common geographic classification (Principle 3 of the GSGF), the grid type geographies complement administrative and statistical geographies.

521. Gridded data can be both a rich source of information and a consistent geography for disseminating and integrating information. Recent global efforts have culminated in the development of a Discrete Global Grid Systems (DGGS) standard which has been developed under the auspices of the Open Geospatial Consortium (OGC). This system offers further options in the use of grids within the context of the principle of common geographies and in geospatially-enabled statistics.

522. To evaluate the comparability of census statistics between different areas, it is necessary to harmonise the data according to international standards. It is therefore important to accompany statistics for the grid with quality metadata.

523. More information regarding the confidentiality and disclosure implications of using grid square and other small area geographies are explained in section ‎7.1.

524. Depending on the geographical frame and the technology used, geocoding may need to be done at the same time as data collection using inputs from GPS systems and putting a mark on a map. The management of the data providers (register owners) is necessary for ensuring that the relationship between the statistical office and data providers remains positive, and for recording and responding to comments, queries and complaints.

525. In addition to the traditional collection of data by enumerators and supplementing this data by geocoding, register, administrative data and even big data are increasingly used. Therefore, it is necessary to ensure the quality of address data, as they are crucial for the correct geocoding of individual data from alternative sources. Geocoded data should be treated in the same way, regardless of its source. This applies especially to censuses using data from registers and administrative sources.

526. Statistical offices should understand GIS as an important tool for spatial analysis and dissemination of census data. Interactive maps, geoportals, and user-centred geography construction tools tend to be more direct and simple ways to interpret information than tabular data, especially for the lay public. NSOs provide vital information about current population, demographic events and future trends to policymakers in a range of sectors, such as health care, education, infrastructure planning and natural resource management; and the availability of spatially-referenced census databases is an essential prerequisite to facilitate the use of demographic data.

527. Maps make it easier to present, analyse and disseminate census results after data processing. The cartographic presentation of census results provides a powerful means for visualizing the outcomes of a census. This supports the identification of local patterns of important demographic and social indicators. Maps are thus an integral part of policy analysis in the public and private sectors.

528. Dissemination of census results in the form of map outputs plays an increasingly important role in data dissemination. In this way, maps are effectively used to link statistical data with the geographic area to which they relate. The results of the statistics are thus easier to understand by the general public and not only to a narrow circle of experts.

529. As for thematic maps, the priority indicators for the population and housing census are the total number of population and dwellings. Other important indicators are age, sex and/or gender, education, employment, types of census household, type of dwelling, form of dwelling ownership, number of rooms and others. Creating maps using the same set of indicators allows countries to compare their evolution over time in a meaningful way. Maps are an invaluable aid when comparing results even at a lower than national level. New technologies provide great flexibility in building informative and visually appealing maps.

530. In addition to classic thematic maps, the creation of grids for the spatial representation of data is increasingly used in statistics. When creating grids, it is necessary to follow a uniform coordination system, which is necessary for the correct connection of geodatabases and the creation of correct map outputs.

531. Nowadays, many tools are available for implementing solutions that allow the user to visualize the boundaries of administrative and statistical geometries and map census indicators such as those identified above. These tools must have an intuitive interface and make it possible to navigate interactively on the map and obtain the census statistical value for each territorial unit, including small area statistics.

532. Options for defining distribution criteria, changing the colour palette, or even statistical analysis tools for exporting data are increasingly sought after by users.

533. In addition to the enumeration areas, census maps can define smaller geographical areas. In many countries, these areas are called small statistical areas or census blocks. In situations where it is possible to count the number of housing units or people for each small area, automatic or semi-automatic spatial aggregation processes are used to delineate the enumeration area.

534. The size of the enumeration areas can vary across the territory, depending on various factors, including the type of settlement in the territory (e.g., urban or rural). As a rule, there is no predefined value for small statistical areas or census blocks. The definition of their geometry is usually based on territorial criteria and the definition of functional geographies that allow for convenient spatial analyses. However, concerns about the number of housing units or individuals these areas include must be taken into account, to avoid confidentiality problems when publishing results.

535. The possible use of these small statistical areas to delineate functional geographies makes it easier to understand regional dynamics and define public policies.

536. For production of small area statistics, population and housing census data collected to the level of x, y coordinates will make a foundation for aggregation to the area that is the subject of statistical analysis of certain social, demographic and economic characteristics. Moreover, for many countries, the census data will also provide a unique source of solid information to develop sampling frames.

537. The growing demand for detailed georeferenced statistical data requires the use and promotion of advanced and efficient Statistical Disclosure Control (SDC) methods. The staff of NSOs should have a broad knowledge of various SDC methods and tools, be able to apply them to obtain sufficiently safe and useful data to be released. The knowledge of efficient ways of protecting the privacy of data stakeholders or customers is also important for many data custodians, who collect data to perform their activities and tasks. Users of statistical data (e.g. students, scientists and analysts) should also be aware of the benefits, drawbacks and expected effects of SDC methods. Detailed information about SDC methods is provided in section ‎7.1.2.

538. Census and geospatial data used to create geospatially-enabled statistics should only be released publicly in a manner that complies with privacy and confidentiality legalisation and prevailing community expectations. The ability to identify personal and confidential information may be possible within data when collected and securely stored but should be anonymised to an appropriate degree when disseminated. Disclosure control process is independent from GIS tools. GIS is used only to visualize and analyze data that should be already anonymised before dissemination stage (see section ‎7.2).

539. Cartography and maps in general are a support for carrying out the different stages of censuses and surveys. The maps ensure complete coverage of the territory, support data collection, supervision of the load assigned to field personnel, processing, analysis of results and dissemination of the results.⁴³

540. A permanently updated cartography allows the different stages of the development of a census to be carried out in a more efficient way. In order to ensure cartographic updating that reflects changes in the territory as a continuous process, it is essential to take advantage of technological advances that include the widespread availability of personal computers, handheld computers and personal digital assistants, GPS and GIS software, and low-cost aerial and satellite imagery (including automated processes for the detection of new human settlements on images for updating the base cartography).⁴⁴

541. The updating of the base cartography is the input for the definition of the common geographies of the statistical and geospatial framework, so the identification of the places where there is the greatest growth in the observation units, such as population, housing, and economic establishments, among others, requires substantial resources.

542. Some of the elements that must be updated in the cartography are: roads, administrative division boundaries, reference elements such as accurately named and presented roads and waterways, and landmark features, such as schools, place of worship, post offices, parks and large buildings based on blocks and localities.⁴⁵ The updating of the base maps will result in better precision and completeness of the enumeration areas and, consequently, a more effective collection of information.