COD-PS Standards and Process

Owned by FIS (Unlicensed)
Last updated: Mar 03, 2021 by FIS
5 min read

Standards

The population statistics COD standards and guidance are currently in development at the global level with UNFPA. The minimum standards identified in the current evaluation checklist enable data interoperability and harmonization. 

COD-PS Required Characteristics (minimum requirements in red)

Metadata

  • Source organization must be clearly identified, and acceptable for humanitarian use

  • Population data are endorsed by IM Network

  • The dataset must be able to be shared (at least once at the onset stage of an emergency)

  • P-codes from COD-PS match COD-AB (tables can link 1:1)

Tabular Attributes

  • Field names are clear and understandable

  • Field names used consistently across all Admin layers

  • Population breakdown exists for sex and age for each administrative unit (with sex disaggregation for every age cohort)

  • All values must be integers.

  • All values must be present

  • Sum of data matches the value of higher admin level

  • Data is checked for inconsistencies

  • Data is checked for outliers

  • P-codes are the unique identifiers used in the country (ideally government sourced)

  • P-codes are present and unique for each administrative unit

  • P-code attribution is consistent across all layers

  • P-codes for higher-level administrative units are included in lower levels

  • P-codes and feature names (and therefore feature counts) should conform to established and accepted administrative boundary datasets (COD-AB) - however, it may be that the COD-PS is the more reliable authority.

  • Identification of Datasets

    The primary data source of data should always be the government or in-country institution supporting a country’s mapping activities. Ideally, a partnership with UNFPA should be used rather than direct contact with the government to acquire a dataset.  Alternatives to government sources should be examined if there is nothing available from an official source, if the official source can not be processed to an acceptable level or if the dataset cannot be shared with the humanitarian community. The COD-PS should be identified via the IM Network and processed at the same time as the COD-AB as they need to have a one to one relationship which may impact the dataset chosen. 

    Consider the following criteria when identifying sources for COD-PS:

    • COD-PS should be operationally relevant and useful.

    • What are partners currently using and why?

    • Is the data source authoritative? If not, what is the rationale for identifying it as a COD?

    • Can it be linked (via P-Code or name) to the COD Administrative Boundary (COD-AB)?

    • To what administrative level is the data provided? What is the lowest administrative level is operationally required?

    • Is it age and sex-disaggregated (can an estimation be made)?

    • Are various sources required to create a complete COD-PS?

    • Acquire as much metadata about the datasets and keep a data trail

    • If various sources are available or being used by partners compile these in a table for easy analysis and discussion with partners. (Example of the table)

    Government Sources are the most authoritative sources.

    The most authoritative source for the COD-PS is the national government.  The main reason for this is because humanitarian agencies will be working with the government and likely receiving other CODs / datasets/information from the government and alignment (P-code/name) is more likely to occur, thus helping communication, coordination, and development of a common demographic standard and understanding. The governmental agencies/ministries responsible for Population Statistics is likely different in each country (e.g., Population Census Office, Agency of Land Affairs, Bureau of Statistics, Ministry of Planning, etc.). In some cases, various levels of government may need to be consulted to acquire or understand the various administrative levels in a country.

    Alternatives to Government Sources

    In some situations, government sources may not be available  (e.g., there may be no data for a particular area of the country that is in conflict; the government is not able to share for various reasons, or a COD-AB is required quickly) and alternative sources need to be identified.  The first question to ask if government data is not available: what are operational (including development) partners using? Other considerations when contemplating alternative sources are: why is this source authoritative; will using a non-government source impact the humanitarian community's relationship with the government; is the sharing/distribution of the dataset the biggest challenge and can this be overcome with a "humanitarian only, for emergencies only" agreement required; is the government likely to release data if an emergency does occur (thus requiring the community to quickly analyze, clean, and share a new dataset); what are partners using and why; and will exposing an alternate source push the government to share data?  Some ideas of alternative datasets are below both categories may be acceptable and the rationale for choosing this source should be included in the metadata. 

Data source 

Advantage 

Disadvantage 

Purpose and process 

Data source 

Advantage 

Disadvantage 

Purpose and process 

Landscan 

  •  



 

World pop 

 

 

 

ESRI 

 

 

 

GHS Population Grid 

 

 

 

GPW v4 

 

 

 

Facebook population data 

 

 

 

Other 



 

 

Process

The COD-PS involves the IMWG and follows the normal COD cycle.

Steps

  1.  Identify (possibly additional) partners in your IM Network who are 'experts' in population statistics/demographic data. UNFPA should be included as defined by Joint letter regarding cooperation between UNFPA and OCHA on CODs

  2. Identify population datasets for a country that are available at different administrative levels and ideally age and sex-disaggregated. They need to match the COD-AB and should be identified and processed at the same time.

  3. If datasets require processing to improve them (e.g. estimations, SADD, etc) the methodology should agree to by the IM network and include in the metadata. This should be done during preparedness activities as it is expected that within 48 hours of an event triggering a humanitarian emergency,

  4. Gain endorsement by the IM Network that the dataset is the COD-PS

  5. Gain endorsement by the HC/RC  of the COD-PS

  6. The COD-PS dataset and metadata should be widely shared through national, regional, and global channels (HDX) to support data compatibility pertaining to the emergency.

IM Network

The IM Network may already include COD-PS subject matter experts but you may want to consider others who are not typically part of the group.  You may also consider creating a specific working group to focus on this particular dataset and report to the larger network after each phase of the COD cycle.  Relationships are a key part of the COD cycle and just as important and the datasets.  Questions to ask yourself when considering the IM Network is for the development of COD-PS: 

  1. Who can provide population statistics datasets? 

  2. Who is using demographic data in-country, regionally or globally? 

  3. Who can help with providing answers to specific questions about the dataset? 

  4. Who can help process the data?  Who can help maintain the dataset over time? 

  5. Who could be responsible for the data post the event?

In addition to the IMWG / IM network members, these particular partners should be consulted: 

  • Government: the ideal data source; can help develop estimates; can help clarify links between COD-PS and COD - AB

  • UNFPA: work with the government on population data

  • UNHCR: to provide refugee population data

  • IOM: to provide IDP population data

  • Others as required

Considerations

  • Leaving no one off the map A gridded population data report from a  wide range of experts: Key messages: 1) census data are still important, gridded population data are not a substitute for census data; 2) gridded population data are not error-free; 3) data users should consider a number of factors when selecting an appropriate dataset for their particular needs; 4) more validation work is needed to compare gridded population data estimates against authoritative data on population location.

Resources