The first step to publishing open data is to make available all the data you collect once privacy and security measures are addressed. ODIN looks on all official government websites for data publish on 60+ indicators. In addition, most indicators require some level of categorical disaggregation (example: sex disaggregation) to be considered. The table below shows the indicators that were not found, with any of the categorical disaggregations listed, on any website. Each category has a certain number of indicators and disaggregations required for full credit. Not all indicators in ODIN are relevant to all countries and scoring rules account for this. For example, in Resource Use only 3 out of 4 indicators are required. If you currently publish one, the other three will appear below, but only two additional need to be published for full credit.

See the Methodology Guide for more information.

It is important to publish recent time-series data, if collected, and not only historical time-series data. The table below shows the indicators in ODIN that were found published and disaggregated, but did not have a single timeseries with at least 3 of the last 5 years

It is important to publish historical timeseries data, if collected, so progress can be assessed over time. The table below shows the indicators in ODIN that were found published and disaggregated, but did not have a single timeseries with at least 6 of the last 10 years. At least 6 of the last 10 years must be published in a timeseries dataset for each indicator to get full credit on data coverage.

Geographic disaggregation increases the value and usefulness of data and enables users to make meaningful comparisons across regions. ODIN assesses geographic disaggregation using two levels- subnational level 1 and subnational level 2. In most countries, subnational level 1 matches their first administrative level/NUTS 2 divisions such as provinces, states, regions, and districts. Subnational level 2 is defined as as geographic division more granular than level 1.

In ODIN, many economic and environmental indicators are not scored on their availability of subnational level 1 or 2 data. Below are a list of indicators, where subnational level 1 or 2 data are required for full credit, but could not be found.

Each website where government data are published should also have a clearly visible data license that meets openness standards, such as Creative Commons BY 4.0. If data cannot be found on the national statistics office's website or portal, other official sources are sought. Any dataset assessed in ODIN that was published on a website without a license or without an open license is shown below.

License are categorized as the following:

1. Not Available: No license was found on the website.
2. Not Open: License published but contains highly restrictive clauses. The most common reason licenses receive this classification is because they forbid commercial use of data.
3. Semi-Restrictive: License doesn't meet all openness standards, but has some conditions that impact its openness.
4. Open: License meets all openness standards.

To get help on how to improve your license, please email us.

All open data should be published in at least one format that is machine readable and nonpropritary. Ideally, the format would be both. Machine readable file formats allow users to easily process data using a computer. Nonproprietary formats are important because they allow users to access data without requiring the use of a costly and exclusive software that may prevent some users from accessing the data. Both are crucial for enabling open access to data. Common file formats are both machine readable and nonproproetary include CSV and XLSX. PDF files, though commonly used, are a non-proprietary format, but not machine readable.

The table below shows all the datasets assessed in ODIN that were not published in at least one machine-readable and non-proprietary format.

File formats that have an "r" next to them are machine readable and those that have an "np" are nonproprietary.

Metadata is crucial for data openness because it provides essential context, such as the source, definitions, classifications used, making it easier to understand, access, and use effectively. It ensures transparency, improves data discoverability, and supports reproducibility in research and analysis. For all datasets assessed in ODIN, reference metadata are reviewed for some basic components common in all recommended metadata standards, such as SDMX, DCAT, and Dublin Core. Those components include: 1) Source agency responsible for collecting the data, 2) Definition of indicator and key terms, and 3) Date that data was uploaded online. Below is a table of the ODIN indicators for which data with incomplete metadata was found (not all three components were found). To be considered, reference metadata must appear in the same file as the data, near the dataset in a separate file or clearly visible, or in a separate section of the same website that is easy found and named appropiately.