Suggested no-key importers: UNICEF, UNHCR, PIP, ILOSTAT, WHO GHO, UNSD SDG

[waindewane]

Summary

I would like to suggest a small set of additional no-private-key importers for bblocks-data-importers that seem well aligned with the package's development-data scope.

The current importer list already covers IMF WEO, IMF DSA, World Bank, World Bank IDS, WHO GHED, UNAIDS, UNDP HDR, and WFP food-security/inflation data, with CEPII BACI listed as upcoming. The docs also point to standalone unesco-reader and oda-reader packages. The sources below would complement those without duplicating the currently listed importers.

Suggested importers

Source	Why it seems useful	Public API shape	Suggested priority
UNICEF Data Warehouse / UNICEF SDMX	Child health, nutrition, WASH, education, child poverty, and child-related SDG data. This is a strong development-data fit and only partially overlaps with WDI/UNData.	SDMX 2.1 XML dataflows and observations. Example dataflow endpoint: https://sdmx.data.unicef.org/ws/public/sdmxapi/rest/dataflow/UNICEF/all/latest?detail=allstubs; example data endpoint: https://sdmx.data.unicef.org/ws/public/sdmxapi/rest/data/UNICEF,CME,1.0/all?startPeriod=2022&endPeriod=2022.	High
UNHCR Refugee Statistics API	Refugees, asylum seekers, IDPs, stateless persons, demographics, asylum applications, and solutions. Useful for humanitarian, migration, fragility, and displacement analysis.	Paginated JSON under https://api.unhcr.org/population/v1/. Example: https://api.unhcr.org/population/v1/population/?yearFrom=2022&yearTo=2022&limit=5.	High
World Bank Poverty and Inequality Platform (PIP)	Dedicated poverty/inequality estimates and regional aggregates, distinct from generic WDI poverty indicators. Useful for poverty-line, survey-year, and regional/global aggregate workflows.	JSON under https://api.worldbank.org/pip/v1/. Example: https://api.worldbank.org/pip/v1/pip-grp?country=all&year=2020&povline=2.15.	High
ILOSTAT	Labor-market indicators such as unemployment, employment, wages, informality, working poverty, and SDG labor indicators.	Public ILOSTAT APIs/SDMX endpoints; no private key required. Example tested indicator: UNE_DEAP_SEX_AGE_RT for unemployment by sex and age.	High
WHO Global Health Observatory (GHO)	Broader WHO health data beyond GHED: mortality, life expectancy, disease, risk factors, UHC, and health-system indicators.	OData JSON under https://ghoapi.azureedge.net/api/. Example: https://ghoapi.azureedge.net/api/WHOSIS_000001?$filter=SpatialDim%20eq%20%27USA%27%20and%20TimeDim%20ge%202020%20and%20TimeDim%20le%202020.	Medium/high
UNSD Global SDG Indicators API	Official SDG indicator series with disaggregation metadata. Some overlap exists with UNData-style access, but the official SDG API has useful series, attribute, and dimension metadata.	JSON under https://unstats.un.org/SDGAPI/v1/sdg/. Example: https://unstats.un.org/SDGAPI/v1/sdg/Series/Data?seriesCode=SI_POV_DAY1&areaCode=840&page=1&pageSize=1000.	Medium

Validation

I prototyped small no-key search/download paths for each source in a local economics-data workflow and verified that the endpoints return usable rows:

• UNICEF CME, 2022: 5 observation rows returned from SDMX XML.
• UNHCR population, 2022: 1 aggregate row returned.
• PIP pip-grp, 2020, poverty line 2.15: 5 aggregate rows returned.
• ILOSTAT UNE_DEAP_SEX_AGE_RT, USA, 2023: 84 rows returned.
• WHO GHO WHOSIS_000001, USA, 2020: 3 rows returned.
• UNSD SDG SI_POV_DAY1, area code 840, 2020: 1 row returned.

All of these worked without private API keys.

Possible API design

The existing bblocks-data-importers interface could probably expose these as separate source classes/functions rather than hiding them behind broader aggregators:

• UNICEF or UNICEFDataWarehouse
• UNHCR
• PIP or WorldBankPIP
• ILOSTAT
• WHOGHO
• UNSDSDG

For consistency with existing importers, each would ideally provide:

• source metadata/search/list methods where the source exposes catalog metadata;
• a get_data(...) style method returning tidy data frames;
• source URL and access-date metadata;
• documented parameter examples for common use cases;
• conservative pagination/row-limit defaults for large APIs.

Happy to share endpoint notes or prototype logic if useful.